It is not uncommon for students to have fixed mindsets about statistics. Many people don’t realize how much they love data until they take their first statistics class. This was the case for Eric J. Daza, a biostatistician and health data scientist with more than two decades of experience who, throughout his educational and professional experiences, also learned statistics can be a philosophical framework for life.
Among his accomplishments, Daza has contributed to the evolution of personalized health data analysis. His work at Evidation Health and Stats-of-1 demonstrates how data scientists and statisticians can change the world for the better.
Daza is also at the forefront of the American Statistical Association’s efforts to improve diversity and inclusion practices across the profession as the professional development chair of the ASA’s Justice, Equity, Diversity, and Inclusion (JEDI) Outreach Group.
We caught up with Daza to learn about his passion for biostatistics; the difference his work makes in health care; and his work advocating for justice, diversity, equity, and inclusion in the field. Take a look at what he had to say.
Did you always want to study statistics?
I was a biology major and didn’t discover statistics until the end of my undergrad years. I loved it, but it was too late to change my major because I was graduating. So, I ended up getting a master’s in statistics and then, later, a doctorate in it. My journey to statistics was pretty much by accident.
I actually sucked at statistics as an undergrad! Throughout college, I struggled not only with statistics but also my major, biology. This was more a reflection of my inability to stay focused than a lack of interest.
I’m pretty sure I ended up taking Introduction to Statistics at least three times, until it finally stuck years later during my biostatistics doctoral program. Biostatistics, by the way, is a misnomer (in this biologist’s opinion), because it involves very little biology and a lot of math!
How has statistics informed your work?
I work as a biostatistician and health data scientist at Evidation, a digital health company. We’re on a mission “to create new ways to measure and improve health in everyday life.”
I code almost exclusively in R, using Jupyter notebooks running on a cluster. I like learning Python but really only do so whenever I have to modify another analyst’s Python code to complete a project. This method of learning works well for me. I also code in SQL from time to time.
For task coordination, I use project and task management enterprise software. These tools help me track progress on action items for myself and my teammates, collaborate and communicate in a way that is time-stamped and auditable, and tighten up project management and coordination.
I also rely on standard operating procedures for streamlining analysis planning and execution, data and code review, and data and report delivery. I use company templates to write study protocols, data management plans, and statistical analysis plans—just like I did when I worked as a biostatistician in the pharmaceutical industry.
When I am challenged as a statistician to balance statistical rigor with time management, I always try to enforce good statistical hygiene to prevent p-hacking, hypothesizing after the results are known (HARKing), overfitting, overgeneralizing, or being too statistically confident with too small confidence intervals.
At my independent newsletter, Stats-of-1, my co-editors and I are on a mission to “promote the expanded use of n-of-1 trials, single-case designs, and other individual-focused (personalized/precision) statistical approaches in health and medicine.” It’s been getting attention; Stats-of-1 was recognized by both Forbes and Fortune in 2022! This work also resonates with what we do at Evidation, so I’m lucky to have this really nice synergy between my day job and ‘extracurricular’ work.
Statistics has really helped structure how I think of ideas going into an analysis and how to be really careful with drawing inference—being really careful about not being too confident in what you find. That’s really, for me, what statistics brings. It’s a framework for scientific inquiry. It’s not just numbers and plug-and-chug.
What do you wish people knew about statistics?
If you wish to pursue a career in statistics or data science, you have to be a generalist, rather than a specialist. Unless you are hired into a specific technical role such as a biostatistician or machine learning engineer, you should seek to develop general statistics and/or data science skills.
As a generalist, you will be able to manage several client projects that may include vague requests by dissecting and helping to clarify, so everyone can focus on the client’s exact underlying goals. By acting as a translator for your clients, you can learn how to present technical language in a comprehensive way to both technical and nontechnical audiences.
Nonstatisticians—myself included when I don’t have my hat on—think statisticians are about the numbers. They see us on the news—and old newsreels will have us typing on a calculator. We’re actually a lot more about the symbols that represent the numbers and how those symbols interact to help explain something about the world. So we’re not actually number crunchers. We’re more like symbol crunchers!
Tell us about the ASA JEDI program.
JEDI—which stands for Justice, Equity, Diversity, and Inclusion—is an ASA outreach group. We’ve been around for about two years, but the group started forming before that.
Our main goal is to foster those principles in the professions of statistics and data science. On a day-to-day basis, we help groups already doing deep work in different segments of the ASA. For example, the Committee on Minorities in Statistics connects to other groups doing similar work to JEDI but in a different area of our association.
The second line of projects we do is developing educational or didactic videos and training for folks such as department chairs and industry leaders. These resources provide information about how to equitably write promotion letters, advocate for employees seeking promotion, and write referral letters for employees applying to a different company.
I’m both a member of JEDI and the ASA. Being a member of the ASA helps me keep up with developments in the statistics profession. It also helps me grow my network of colleagues. An ASA membership also provides discounts for ASA events like the Joint Statistical Meetings, so there are financial benefits.
That said, any organization requiring a paid membership to receive benefits needs to be aware of how it may be systematically excluding groups of people from those benefits. This includes opportunities to network, which are so important to professional development.
What advice would you give to future statisticians and data scientists?
If you are a young professional about to enter the professions of statistics and data science, you should know the following two things:
- We are guardians of the scientific method.
- Being a statistician is awesome.
We are responsible for recognizing sources of uncertainty in data collection and analysis and quantifying their impact on the strength of scientific conclusions to set our research collaborators up for success.
As the field of data science continues to grow, it is crucial that we clearly explain our epistemological role to nonstatistician colleagues. We must clearly demonstrate the value our role brings to data science—not as competitors, but as partners, collaborators, and practitioners.
Remember, your job is to help your collaborators succeed by understanding and managing their expectations around the type and strength of their evidence—what their data, assumptions, and hypotheses (including models) together can and cannot say.
See Daza in action in this ThisIsStatistics interview.
Leave a Reply